Hi,

Posted to TBI board and guys from there thought that maybe could someone from here answer better. So sorry about double posting.

The problem - I could not pass fuel emission test as they found the lambda constant being too high (1.3 on 2000 rpm). CO, HC, all other things measured were OK. Testing person swore that the engine should get more air than needed and running lean. But the fuel consumption and fuel smell from exhaust seemed to indicate more a rich mixture for me. I went to diagnostics and the guy there said that the ECM was bad (in closed loop the mixture was way off and rpm changed periodically from 400 to 800 rpms). I changed the ECM and decided to hang computer to the car myself. It idles fine now - rpm is very stable, but it runs rich again. O2 reading changes from 200 mV to 900 mV, BLM were 117 and INT changed from 122-134. All other readings seemed to be fine for me.

For some reason I checked the injector pulses with oscilloscope and find something weird. I sketched the normal pulses, as they should be by LM1949 datasheet (up) and the real pulses from my car (down). Real pulses are shrinked a little bit (so the timeline is not 1:1), there is 2 pulses on upper figure and 4 on the other:



Any ideas would be appreciated, as my brain cells are pretty much done at the moment

 

I uploaded a data that I logged yesterday too (few minutes usual city traffic), maybe someone could find something from there: http://lepo.it.da.ut.ee/~rnahkur/Staff/logi.xls